Cover Image
close this bookReversing the Spiral - The Population, Agriculture, and Environment Nexus in Sub-Saharan Africa (WB, 1994, 320 p.)
close this folder3. The demographic dimension
View the documentThe lagging demographic transition
View the documentFertility and agriculture: Part of the Nexus?
View the documentNotes
View the documentAppendix to chapter 3

The lagging demographic transition

Sub-Saharan Africa lags behind other regions in its demographic transition The total fertility rate (TFR) for SSA es a whole has remained virtually unchanged at about 6.4 to 6 6 for the past twenty-five years (Table A-2) This is significantly higher than in other countries with similar levels of income, life expectancy, female education, and contraceptive prevalence. In a number of countries of Sub-Saharan Africa fertility in fact has risen (in large part due to significant success in treating diseases that cause infertility), while it has declined elsewhere in the developing world.

Recent statistics, collected through nationally representative sample surveys carried out between 1986 and 1989 under the Demographic and Health Surveys (DHS) Program,¹ appear to signal' however, that several countries are at or near a critical demographic turning point (Table A8).² In Botswana, the TFR fell from 6.9 in the mid-1960s to 4 7 in 1989, and in Zimbabwe it dropped from 8.0 to 5.3 over the same period. In Kenya, the TFR declined from 8.2 in 1977/78 to 7.7 in 1984 and to 65 in 1989 (Kelley and Nobbe 1990:33.). Encouraging, too, are the date from Nigeria which indicate a TFR of 5.7 in 1990, compared with 6 9 in 1965 In Cd'Ivoire Ghana, Mozambique, and Sudan, fertility also appears to have begun a secular decline (Table A-2).³

Life expectancy in Sub-Saharan Africa has risen from an average of 43 years in 1965 to 51 years in 1990 (Table A-1) In eighteen countries, average life expectancy today is 53 years or more Mainly due to the decline in mortality rates, population growth has accelerated from an average of 2.7 percent a year for 1965-1980 to about 3.1 percent a year at present (Table A-2). And, given the age structure of SSA populations, the momentum for continued growth is already built in. Even if the TFR were to drop immediately to the replacement level of 2 2 births per woman, it would take almost a hundred years before the population would cease growing. By then it would be 80 to 100 percent larger than it is today.

Figure 3-1 Total Population of Sub-Saharan Africa, 1960-1990

The high fertility rates, and the marriage, reproductive, and contraceptive behavior patterns that underlie these, arise in part from the fact at most women live in rural areas and have little or no education, few opportunities; outside their traditional roles, and limited legal rights. Childbearing enhances their status, and most women marry and begin having children early and continue to have them throughout their fecund years. But the comparative lack of urbanization and education does not explain everything: data available from the World Fertility Survey (WFS) for 1978-1982 indicate that urban and rural, educated and uneducated women in Sub-Saharan Africa have and want more children than their counterparts elsewhere.4

Women in Sub-Saharan Africa marry early: WFS data for the early 1980s show that, on average, 40 percent of all women aged 15-19 and 75 percent of those age 20-24 were or had been married (Cachrane and Farid 1989). Early female marriage increases the number of fecund years a women spends in union and therefore tends to exert upward pressure on the TFR. Even urban women in Sub-Saharan Africa marry earlier than rural women in North Africa and Asia (Cochrane and Farid 1989).

Contraceptive use in Sub-Saharan Africa is far below that in other regions (Tables A-5 and A 6). It is associated primarily with the desire for child spacing and only secondarily with the wish to limit family size (Table A-6). Use of efficient contraceptive methods generally increases with urbanization.5 Low contraceptive use is due in pert to poor knowledge. There are wide differences among countries, but on average, only about half of all women in SSA had, by the early 1980s, heard of a way (either efficient or inefficient) to prevent pregnancy. This compared with rates of 85 to 95 percent in other regions (Cochrane and Farid 1989). By the late 1980s, the DHS surveys showed measurable increases in the percentage of women who had knowledge of modern contraceptive methods: in ten of the twelve SSA countries surveyed and for which data are available so far, between 64 and 98 percent of currently married women aged 15-49 knew of at least one modern contraceptive method (the exceptions were Mali and Nigeria, with only 29 percent and 41 percent, respectively). The DHS data on contraceptive prevalence rates (CPRs) indicate, however, the difficult task ahead: only between 1 and 6 percent of these married women were currently using a modern contraceptive method, and the percentage of married women using any contraceptive method ranged only between 3 and 13 percent. The exceptions, with significantly higher CPRs, are Botswana, Kenya, and Zimbabwe (Table A8).

Box 3-1 Population Projections for Sub-Saharan Africa

Recent Bank projections assume that the TFR for Sub-Saharan Africa (including South Africa) will decline from 6.5 percent in the period 19851990 to about 3.25 percent by 2030. This implies average annual population growth rates of just over 3 percent in the 1990s, 2.9 percent in the following decade, and close to 2.6 percent in the decade thereafter. In this scenario, the SSA population would exceed 1,000 million by the year 2012 By the turn of the present century, SSA would be second only to Asia in terms of total population. Aggregate population growth will remain above 2 percent per year at least until 2025, by which time the population of SubSaharan Africa, at 1,378 million, would be 2.6 times that of today. The net reproduction rate (NRR) will decline to 1 only by 2060. And Sub-Saharan Africa would reach a hypothetical stationary population of over 3,100 million only some time after the year 2150 (Stephens and others 1991).

Among all groups of women, desired fertility is far higher in SubSaharan Africa than elsewhere. However, the WFS data analyzed by Cochrane and Farid also showed that (a) younger women desire fewer children than do older women; (b) urban women want fewer children than do rural women (although urban residence has not yet become a strong fertility depressant—the rural-urban differentials being smaller in SSA than in other regions); and (c) educated women want far fewer children than do uneducated ones

Box 3-2 Contraceptive Prevalence Among Women: Sub-Saharan Africa vs. Other Regions

Data for the early 1980s indicate that the percentage of women in SubSaharan Africa who had "ever used" any contraceptive methods varied widely-from 2 percent in Mauritania to 74 percent in Cd'Ivoire. The average for the ten SSA countries covered in the World Fertility Survey (WFS) was 26 percent, compared with 40 percent in both North Africa and Asia and 62 percent in Latin America. The regional comparison revealed far greater differences when only "efficient" methods were considered: only 6 percent of women in SSA had ever used these, compared with 32 to 50 percent of women in other regions.

Based on reported "current users," contraceptive prevalence in SSA was very low indeed, of both efficient and inefficient methods. In six of the ten SSA countries surveyed, fewer that 1 percent of all women were current users of modern methods. The difference between "currently using" and "ever used" was much greater in SSA than elsewhere, probably reflecting the comparatively much greater use in SSA of contraceptive practices for birth spacing, rather than for limiting family size (Cochrane and Farid 1989).

In Sub-Saharan Africa, as elsewhere, women's education affects fertility preferences, use of modern contraceptive methods, and fertility. Cochrane and Farid found that:

· There are considerable differences in desired family size among countries (see also Table A-8), but with increasing maternal education there is both a decline and a clear convergence across countries.

· Current use of any contraceptive method was only 4 percent among the least educated (compared with 19 to 34 percent in other regions), but 19 percent among the most educated (compared with 43 to 56 percent in the other regions).

· Although current use of contraceptive practices among the most educated women in Sub-Saharan Africa was only about the same as among the least educated in North Africa and well below the least educated in Asia and Latin America, even this low rate was sufficient to lower the TFR to about 5 for women with seven or more years of schooling.

Fertility rises with a few years of schooling, but then declines (as in other regions). But the effect of maternal education on fertility has been less pronounced, to date, in Sub-Saharan Africa than elsewhere.

The very high infant and child mortality rates (Table A-3) prevent achieving desired, or target, fertility levels—and this helps explain the low CPRs (Tables A-5 and A-6). Contraceptive use increases as the number of living children increases. Although infant and child mortality have declined over the past two decades (in some countries substantially), they remain much higher than in other regions (albeit with considerable differences among countries).6 Higher child survival rates reduce the need to replace children who have died or to have more children to insure against the likelihood of future deaths. Infant mortality rates are well below the SSA average of 107 in Botswana (36), Kenya (67), and Zimbabwe (48) . The same is true for child mortality rates, where the SSA average is 177: Botswana (403, Kenya (105), and Zimbabwe (57).

each of these countries, the TFR has begun to show a decline, signalling the onset of the demographic transition (Table A-8).

Infant mortality is highest in rural areas, and children born to young mothers are at greater risk. Infant mortality is also higher for first-born children and for those born seventh or later. Children's survival chances are greater if the interval from the previous pregnancy is longer (maternal attrition, lower risk of low birth weight' maternal attention). Infant and child mortality decrease consistently with the mothers' education. Urbanrural differences in infant and child mortality are significant and somewhat larger in Sub-Saharan Africa than in other regions.

Prolonged and near universal breastfeeding has been the main factor keeping fertility below a biological maximum in most SSA counties. The duration of breastfeeding is generally shorter in urban than in rural areas, but it does not decline as rapidly with mother's educational levels as in other regions. The most educated women in Sub-Saharan Africa breastfeed considerably longer than those in Latin America and Asia. Breastfeeding has important positive effects on child health and, indirectly, via reduced infant mortality, on fertility decisions. It also affects birth spacing—and thereby maternal health, infant health, and fertility. In this respect, postpartum infecundity is far more important in Sub-Saharan Africa than elsewhere, accounting for 59 percent of the reduction in fertility from the biological maximum. On average, fertility in SSA is only 67 percent of what it would be in the absence of breastfeeding.

Nevertheless, fertility patterns do not seem to be fully explained by the proximate determinants of marriage, postpartum infecundity (breastfeeding), and contraceptive use. Abortion, sterility, subfecundity, and spousal separation appear to suppress the "maximum" fertility below that observed in other regions This suggests the need for more research on other determinants of fertility to understand current levels and probable future trends (Cochrane and Farid 1989).

Figure 3-2 Infant Mortality Rate, 1960-1990

Fertility and agriculture: Part of the Nexus?

Fertility is highest in rural areas—reflecting economic and sociocultural factors which affect fertility aspirations. Traditional lineage and kinship systems, gender roles, and intergenerational relations contain strong pronatalist forces, and women's fertility usually is a major determinant of their status. Extended families, where the costs of high fertility are only partly borne by the couple making the fertility decision, tend to encourage high fertility. In most of rural SSA, labor is not readily available for hire and must be mobilized from within the household or through social or kinship arrangements specific to the community. For men, polygamy (or polygyny) is a widely practiced way of securing the labor of women and their children—but even women may welcome co-wives as co-workers (e.g., Netting 1993:89). Polygamous men generally have more children than monogamous men, while women in polygamous marriages tend to have fewer children than those in monogamous marriages (Bongaarts and others 1990:135-136).

Women may recognize far more readily than men the costs of high fertility to their own and their children's health. This may be particularly prevalent in polygamous unions where each woman is responsible for her own children. The costs of children are lower to men than to women, yet the value of child labor may be higher to the mothers than to the fathers—except in communities where fathers have and assert priority rights to their children's labor. For women, the labor of their children is often the only means of securing adequate labor to cope with their many responsibilities. (In many communities, women try to ease their peak labor constraints by participating in various forms or kinship- or community-based work group and labor exchange arrangements.) As water and woodfuels become more scarce and the time required to obtain them increases, the need increases for children to help with the mothers' growing workload associated with these survival activities. Child labor is also increasingly needed to compensate for declining male labor in food crop production, particularly in poor families that cannot hire wage labor. This may contribute to the persistence of high fertility rates.7 In much of SSA, men and women cultivate different crops on separate plots, and women's farming systems depend heavily on female and child labor. Most women marry at an early age and often considerably older men. Coupled with the high rates of divorce/separation and the fact that in most African societies women can gain access to critical assets (such as land) and public services only through male relatives, this may increase women's willingness to bear many children so as to have sons to turn to when husbands leave or die. The desired number of children is considerably higher among rural women in Sub-Saharan Africa than among their counterparts in any other region of the world. And in no other region of the world do women play as significant a role in agriculture as in Sub-Saharan Africa.

The characteristics of most traditional land tenure systems may also bear upon fertility decisions—but more research is needed to establish this link. Where access to land for farming is granted to all members of a community, this may be a disincentive to fertility control. Where the amount of land allocated is based on the ability to cultivate it, this ability -- under the low-resource farming conditions prevailing in most of Sub-Saharan Africa — is primarily determined by the ability to mobilize labor. In most cases, this means family labor—more specifically, female and child labor. Indeed, a number of field studies report this to be an important incentive to increase family size through such means as polygamy and pressure on women to have many children.

Among groups with matrilineal descent and inheritance traditions, further complications may arise because land use rights are not passed on from fathers to their children, but to uterine relatives (in most cases males). This weakens the link between land availability and land resource management on the one hand and demand for fewer children on the other. It also weakens men's incentives to invest in maintaining the fertility of the land they farm.8 Fathers may see little point in preserving farm land in good condition beyond their own lifetime or in having few children so as to pass on a viable farm unit to each of them. Women, conversely, may face social pressure to bear many children so as to increase the number of future claimants to land resources who belong to their lineage.

The implication derived from the above is that most rural Africans attach high economic value to having large numbers of children. Larger families appear to fare better economically than small families. Children contribute labor in cropping, livestock tending, fishing, water and fuelwood fetching, and child rearing. The available evidence, although imperfect, suggests that high demand for children may be partly the result of the historic abundance of land and the shortage of labor, combined with high infant, child, and overall mortality rates and high food insecurity. Maintaining high fertility is the rational response of people who seek to ensure adequate family labor and the survival of children who would support them in old age. For men in particular, polygyny makes good sense in this situation because it increases the supply of female and child labor and improves the prospects for security in old age. The widespread practice of payment of a bride price (instead of the woman's family providing a dowry) reflects this reality where women are wanted for their labor and their ability to beer many children. Early female marriage, common in Africa, also increases the prospects for multiple childbirths.

Various other trends also tend to keep the TFRs high. As forest resources, water availability, and soil fertility decline, farmers and pastoralists obtain less product per hectare The main resource available to them to increase production is family labor, which permits increasing the extent of the land farmed. It also makes it easier to diversify the sources of family income with more seasonal or full-time off-farm employment. Hence, agricultural stagnation and environmental degradation, in resource poor situations characteristic of most of SubSaharan Africa, provide an economic incentive—and often a survival strategy —to maintain large families. These factors also provide an incentive to keep children out of school to work on the parental farm or with the family's livestock.

This situation is exacerbated by the specific and important responsibilities placed on women in most farming systems of SubSaharan Africa. Women are often responsible for food cropping, and almost always for fuelwood and water provision (Chapter 5). As soil fertility declines and distances to fuelwood and water sources increase, many rural women are faced with the situation that the only resource that can be increased to meet the increasing need for labor is child labor. More labor substitutes for reduced soil fertility and compensates for the greater difficulty in obtaining fuel and water. This then completes a vicious circle in which population growth, combined with traditional farming practices, contributes to environmental degradation, in turn contributing to further agricultural stagnation and to the persistence of high rates of population growth.

These hypotheses are consistent with statistical tests (see the Appendix to Chapter 3) which show that, other things being equal, TFRs are highest in those SSA countries that have the most cultivated land per capita. Similarly, TFRs are highest in countries with the highest infant mortality rates, lowest level of female education, lowest urbanization, and greatest degree of land degradation. This suggests that demand for children as well as TFRs will decline over time—even without an active population policy—as population density on cultivated land increases, and if female school enrollment rates rise, infant mortality declines, urbanization increases, and environmental degradation is minimized.

However, changes in these determinants of the demand for children are coming about only slowly. Analysis of available cross-country data suggests a considerable degree of inertia in fertility rates as well as the presence of many other factors that influence fertility rates but for which data are not available. Cultural elements appear to be very important; Commenting on the findings obtained from the analysis of the WFS data collected in the late 1970s and early 1980s, one of the program leaders stated that the onset of the demographic transition "appears to be determined more by ill-understood cultural factors than by any objectively ascertainable development indicators" (GilIe 1985:279). These cultural determinants are likely to change only slowly, even though many of the factors that help shape culture are changing. Fertility rates will decline, but only slowly, and only if infant mortality declines lines and environmental degradation is arrested. But progress in these two critical areas is occurring too slowly to compensate for the enormous difference between the current rates of growth of population and of agricultural production.

Nonetheless, rising population pressure on cultivated land, declining infant mortality rates and improvements in female education are stimulating demand for family planning services. Much of this demand remains, at present, unmet (Table A-6).


1. the DHS Program, a follow-up to the World Fertility Survey (WFS), is a nineyear program to assist developing countries in implementing fifty-nine demographic and health surveys.

2. The TFR has declined most dramatically in Mauritius, falling from 4.8 in the mid-1960s to 13 by 1990 (Table A-2).

3. Detailed analysis of the DHS data is still in progress, and this study could therefore not yet draw fully on the information collected under the DHS Program. Note also that the data in Table A-2, which represent the "best estimates" currently available in the World Bank's demographic statistical data base, do not in all cases fully reflect the most recent survey findings obtained under the DHS Program.

4. Under the WFS, national surveys were undertaken in the late 1970s and early 1980s. Using these data, Cochrane and Farid (1989) carried out a comparative analysis to ascertain simiIarities and differences in fertility and underlying causal factors between SSA and other regions; when they undertook this study, WFS date were available for ten SSA countries.

5. Interestingly, when efficient and inefficient methods were considered together, urban use was higher than rural use in only three of the ten countries for which WFS data were available (Lesotho, Nigeria, Sudan). Traditional practices of fertility control such as breastfeeding, might have been abandoned in the course of modernization, while modern methods were not yet adopted widely enough to offset this This explanation is frequently given for the small differentials in fertility across socioeconomic groups, but the available data on breastfeeding practices in Sub-Saharan Africa do not support this conjecture. Breastfeeding does not decline rapidly with increasing education (Cochrane end Farid 1989).

6. The differential is higher in the case of child mortality, due to high mortality in the second and third years of life, following weaning. Toddler and child mortality rates are two to three times greater in SSA than in Latin America and Asia. Toddler and child mortality at all levels of mother's education are higher than in other regions (Cochrane and Farid 1989; see also Table A-3).

7. This is suggested, though not necessarily proved, by the statistical analysis summarized in the Appendix to Chapter 3.

8. Among the matrilineal Akan in Ghana, for example, it is frequently observed that a son employed in an urban job is very reluctant to have remittances he sends to his father invested in improving the father's farm ventures because these investments will, upon the father's death, benefit the son's maternal uncles or cousins, rather than himself.

Appendix to chapter 3

Statistical Analysis to Explain Intercountry Variations in Total Fertility Rates

The available data, for thirty - eight countries, were used to test a number of the country-level findings concerning the determinants of high TFRs. A first set of tests was undertaken using cross country data, looking at relationships in the variation of variables across countries. The results were consistent with the analysis presented in Chapter 3, but questions remained concerning the robustness of the results and the statistical fit. A second set of tests was therefore carried out on a much larger set of data which included both time series and cross-country data. The results are reported here, along with any differences with the findings from the first set of tests. The results of both sets of tests are highly consistent (The data and the methodology used are discussed in the Supplement to this volume.)

For these tests, TFRs are hypothesized to be related to the independent variables as follows: positively to infant mortality (the higher the expected loss of infants, the more births are desired to ensure sufficient survivors); negatively to female school enrollment (better educated women want fewer children); negatively to food security (the greater the food security, the lower the need for children to provide farm labor); positively to cultivable land per person (the more cultivated land per person, the greater the need for family labor to work it); positively to the rate of deforestation (the higher the rate of deforestation, the greater the need for child labor to help fetch wood and water); and negatively to urbanization (urbanization lowers the TFR).

These hypotheses are tested by means of statistical regression with the TFR as the dependent variable. The independent variables are all lagged one year. The lag structure is arbitrary; several were tried, but the statistical fit did not improve. The Supplement presents results of tests for combinations of different countries and different data. Only the sign of the coefficient, the range of t-statistics and the range of significance levels in the various tests are reported below. Since the methodology used does not permit the value of the coefficient to be readily interpreted, it is not reported.

Independent Variables

Coefficient t-statistic

2-Tail Significance Test

Infant mortality rate




Female school enrollment rate




Calorie supply as % of requirement




Hectares cultivated per person




Rate of deforestation




Degree of urbanization




Adjusted R squared = 0.44 to 0.46

The coefficients for female school enrollment' area cultivated per person, infant mortality, and the degree of urbanization are statistically significant at above the 90 percent level, with a 2-tail significance test of 10 percent or less (line 2-tail significance test indicates the probability of the coefficient actually being zero Hence, a 2-tail test of 2.2 percent for the rate of female school enrollment indicates a 22 percent probability that the coefficient is zero or a 97.8 percent probability that it is not zero.) Although the relationship between deforestation and the TFR is positive, as hypothesized, the statistical tests do not suggest significance. The coefficient for calorie supply has the wrong sign and is insignificant.

In the tests with single-year cross-country data, the results were essentiaIly the same, except that deforestation was also significantly related to the TFR, and the coefficient for calorie supply was positive, as hypothesized, but insignificant.

These findings suggest that the TFR is lower as female primary school enrollment is higher The greater the area cultivated per person, the higher the TFR. The higher the infant mortality rate, the higher the TFR. The greater the rate of urbanization, the lower the TFR The positive association between the rate of deforestation and the TFR has ambiguous significance This may be because the rate of deforestation is a poor proxy for the rate of degradation of the rural environment which includes soil and water degradation. Or the hypothesis itself maybe incorrect If further analysis establishes the significance of this relationship, it suggests that greater demand for child labor is associated with environmental deterioration (more labor needed to obtain wood and water and to produce food as the productivity of farm land declines due to deforestation) The relationship between nutrition and the TFR is even more ambiguous, and the hypothesis could not be supported statistically. At very low levels of nutrition, improving calorie intake may increase fertility and, hence, the TFR. Or there may be no relationship Better data are needed to resolve this.